Bill Hoffman wrote:
> Bill Hoffman wrote:
>> Joachim Geiger wrote:
>>> Hello,
>>> this is more or less to document a bug in the cmake version 2.6.2 seen
>>> on AIX version 5.3 when using the option to redefine the fortran
>>> compiler. For example:
&g
is important to note that this bug is gone with version 2.6.3.
Additionally, on my linux systems I did not observe this endless loop
with version 2.6.2.
Since I could not find anything related in the mailing list and on the
web, I thought it best to add it to the mailing list.
Best regards,
Joac
with the above settings of the
cmake-cvs did also run the bootstrapTest positive. Thanks for the help.
I will probably use this one now.
Best regards,
Joachim
--
Joachim Geiger [EMAIL PROTECTED]
Max-Planck-Institute for Plasma Physics,@\\
Euratom Association
fails in the test. The bootstrap
test, however, did not run.
Best regards,
Joachim
--
Joachim Geiger [EMAIL PROTECTED]
Max-Planck-Institute for Plasma Physics,@\\
Euratom Association\\\
Department Stellaratortheorie (ST)\\\__
Wendelsteinstr
o be THE FORTRAN-COMPILER to be used?
Best regards,
Joachim
P.S.: the experimental-make did not yet show-up on the dashboard.
--
Joachim Geiger [EMAIL PROTECTED]
Max-Planck-Institute for Plasma Physics,@\\
Euratom Association\\\
Department Stella
Bill Hoffman wrote:
Joachim Geiger wrote:
Bill Hoffman wrote:
So, does CVS cmake need the changes to curl.h and friends, or is it
just 2.4.8?
No, the cvs-cmake does it alright.
Can you send me a patch?
I don't think there will be a 2.4.9 but I can put the patch on the
branch.
Bill Hoffman wrote:
Joachim Geiger wrote:
I checked out the cvs-version yesterday (cmake-2.7.20080326). This
compiled on the newer without problems. However, running the test
resulted in 3 failures:
97% tests passed, 3 tests failed out of 90
The following tests FAILED:
56
Bill Hoffman wrote:
Joachim Geiger wrote:
Sorry for the delay. You didn't scare me away, but I was trying to see
what might be special about the machines I have access to and about
their setup.
So a first result:
There are two machines, one with AIX-5.1 which has due to compatibility
re
Bill Hoffman wrote:
Joachim Geiger wrote:
//No help, variable specified on the command line.
CMAKE_Fortran_COMPILE_FLAGS=-qrealsize:UNINITIALIZED=8
//Flags for Fortran compiler.
CMAKE_Fortran_FLAGS:STRING=-qrealsize=8
So I did not know what to fix there.
However, in the files flags.make and
Bill Hoffman wrote:
Joachim Geiger wrote:
Hello,
I have the following problem with a fortran compiler flag I want to be
accepted. Compiler is the IBM compiler version 8. The compiler flag I
want to be accepted is -qrealsize=8 .
Using CMAKE_VERBOSE_MAKEFILE=ON I see that the compiler flag
-DCMAKE_Fortran_COMPILE_FLAGS=-qrealsize=8
and also with quotes: ' and ", but the effect is always that it is
translated into the above mentioned one.
An additional note: the AIX-compiler is named f95 and is the same as xlf.
Best regards,
Joachim
--
Joachim Geiger [EMAIL PROTECTED]
Max-Planck-Ins
ptr.hxx",
line 81.3: 1540-1283 (I) "cmsys::auto_ptr::auto_ptr(auto_ptr &)" is not a viable candidate.
I am currently doing some testing on the AIX-5.3-system with some
projects, mixed c and fortran, using cmake-2.7, but I have some problems
to get cmake to accept the fortran co
68 Testing complex ***Failed
39/ 68 Testing complexOneConfig ***Failed
63/ 68 Testing kwsys ***Failed
The behaviour was the same for aix-5.1 and for aix-5.3.
I hope this helps to improve cmake.
Best regards,
Joachim
--
J
13 matches
Mail list logo